CN113393028A - Load prediction method based on data mining technology - Google Patents

Load prediction method based on data mining technology Download PDF

Info

Publication number
CN113393028A
CN113393028A CN202110651314.8A CN202110651314A CN113393028A CN 113393028 A CN113393028 A CN 113393028A CN 202110651314 A CN202110651314 A CN 202110651314A CN 113393028 A CN113393028 A CN 113393028A
Authority
CN
China
Prior art keywords
load
data
feeder
feeder line
correlation
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202110651314.8A
Other languages
Chinese (zh)
Other versions
CN113393028B (en
Inventor
张乐
丁小叶
张敏
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Nantong Power Supply Co of State Grid Jiangsu Electric Power Co Ltd
Original Assignee
Nantong Power Supply Co of State Grid Jiangsu Electric Power Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Nantong Power Supply Co of State Grid Jiangsu Electric Power Co Ltd filed Critical Nantong Power Supply Co of State Grid Jiangsu Electric Power Co Ltd
Priority to CN202110651314.8A priority Critical patent/CN113393028B/en
Publication of CN113393028A publication Critical patent/CN113393028A/en
Application granted granted Critical
Publication of CN113393028B publication Critical patent/CN113393028B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/23—Clustering techniques
    • G06F18/232—Non-hierarchical techniques
    • G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
    • G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
    • G06Q50/06—Electricity, gas or water supply
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
    • Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
    • Y04S10/00—Systems supporting electrical power generation, transmission or distribution
    • Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications

Abstract

The invention discloses a load prediction method based on a data mining technology, which comprises the following steps: (1) obtaining feeder line data and preprocessing; (2) analyzing the load characteristics of various feeder lines by using a clustering method in a data mining technology; (3) analyzing the obtained feeder line loads of each type, and qualifying the electricity utilization characteristics of the industry to which the feeder line loads of each type belong to achieve the purpose of fine analysis; (4) constructing an influence factor set; (5) mining a plurality of factors influencing the load characteristics and each type of feeder load by using correlation analysis and association rule analysis, and researching the association characteristics between the feeder load characteristics and the influence factors thereof; (6) and respectively constructing a neural network prediction model for each type of feeder load and influence factors to complete load prediction. The invention can obtain the load forecasting condition with more scientificity and accuracy, thereby being beneficial to comprehensively investigating the load conditions of distribution and transformation in different areas and being beneficial to the operation management and the dispatching planning of the power grid.

Description

Load prediction method based on data mining technology
Technical Field
The invention belongs to the technical field of electric power, and particularly relates to a load prediction method based on a data mining technology.
Background
The prediction of the power load has great significance for power grid operation planning management and power dispatching plan formulation, and the accurate power load prediction method can realize accurate planning of power energy and is an important guarantee for stable power utilization of users, stable operation of the power grid and stable economic development. In the past, common load prediction methods such as a time series method, a trend extrapolation method, a regression analysis method, a gray model method and the like need to establish an accurate and complex model. The development of big data technology and data mining technology provides a feasible way for the high-efficiency processing of massive load data, and common data mining technologies for load prediction include a support vector machine, a neural network and cluster analysis.
Because the system load is composed of a plurality of electric loads, the electric loads are varied widely, different types of electric loads have self load characteristics and load development rules, and the electric loads can weaken or even counteract the variation rules of certain electric loads when being superposed, so that the regularity of the system load becomes fuzzy, and the real reason of load fluctuation is difficult to accurately position; meanwhile, due to the characteristics of a plurality of load influencing factors, nonlinearity, complexity, hysteresis and the like, the establishment of a relation model between the system load and the plurality of influencing factors in practical application is very difficult. Therefore, the conventional load prediction accuracy is not high.
Disclosure of Invention
The invention aims to overcome the defects of the existing load prediction technology and provides a power load prediction method based on a data mining technology aiming at massive power grid operation data.
The invention specifically relates to a load prediction method based on a data mining technology, which comprises the following steps:
(1) acquiring feeder line data and preprocessing the feeder line data;
(2) analyzing the load characteristics of various feeder lines by using a clustering method in a data mining technology, and analyzing the obtained load of each feeder line;
(3) the power utilization characteristics of the industry to which each type of feeder load belongs are determined according to the load level and the load curve shape of each feeder cluster;
(4) constructing an influence factor set;
(5) analyzing a plurality of factors influencing the load characteristics and each type of feeder line load by using correlation analysis and association rule analysis, and excavating association characteristics between the feeder line load characteristics and the influence factors thereof;
(6) and predicting each feeder line cluster by using the RBF neural network, obtaining a total power load predicted value of the system at the time to be predicted according to the load predicted value of each feeder line cluster at the time to be predicted, and completing load prediction based on the data mining technology.
Further, the step (2) specifically includes the following steps:
21) initializing a clustering center: randomly selecting C clustering centers;
22) calculating the mass center: continuously and iteratively calculating membership degree muijAnd cluster center cjUntil they reach the optimum;
wherein degree of membership
Figure BDA0003111294830000021
Each sample data xiAssign to the nearest cluster center, form C clusters, the objective function is as follows:
Figure BDA0003111294830000022
wherein: m is the number of clusters of the cluster; c is the number of the clustering centers appointed in advance; x is the number ofiIs the ith sample; cjIs the center of the j cluster; mu.sijIs a sample xiMembership to a jth cluster center;
23) continuously correcting the clustering center by an iteration method until the preset target function precision is met, namely when | | | muij (k+1)-μij (k)Stopping iteration if | | < epsilon, otherwise returning to the step 2);
24) and carrying out dimensionless normalization on the classified feeder line data, and analyzing the electricity utilization characteristics according to the load curve.
Further, the set of influence factors of the feeder line load in the step (4) includes economic development indexes, social resident development indexes and climate indexes, the economic development indexes include GDP, a first industrial production total value, a second industrial production total value, a third industrial production total value and an industrial production total value of more than scale, and the social resident development indexes include a total number of people in a region, a total number of domestic production in a per capita, a resident consumption price index and a urbanization rate.
Further, the step (5) specifically includes the following steps:
51) after the influence factor set is constructed, judging the correlation between each influence factor and the clustered feeder load data by adopting a Pearson correlation coefficient analysis method;
52) performing association rule analysis on the strongly relevant factors by adopting a grey association analysis method;
53) and obtaining factors with strong relevance, and taking the factors and the clustered various feeder loads as input to train the neural network.
Further, the step 51) specifically includes the following steps:
1) using the clustered feeder load data as a reference sequence X0Each influencing factor sequence is a comparison sequence Xi；
2) Dimensionless treatment: processing each data into data under the same dimension through normalization processing;
3) and calculating the correlation, wherein the Pearson correlation coefficient calculation formula is as follows:
Figure BDA0003111294830000031
and calculating the correlation strength of each influence factor and feeder line data through a Pearson correlation coefficient, and then keeping the influence factors with the strong correlation.
Further, the step 52) specifically includes the following steps:
1) selecting a reference sequence, namely feeder line data, and comparing the sequences, namely the strong correlation factors after correlation analysis and screening;
2) data transformation: processing and transforming the data to ensure that the grey correlation analysis is carried out under the same dimension;
3) calculating the relevance: the Dun model of a typical gray correlation model is adopted, and the correlation calculation formula is as follows:
Figure BDA0003111294830000032
wherein: xi (X)0(k),Xi(k) Referred to as reference sequence versus comparison sequence
Figure BDA0003111294830000033
Further, in the step (6):
the data flow of the RBF neural network is as follows: training a sample, namely an RBF hidden layer, a weight matrix and an output layer, wherein nonlinear mapping is performed from the input layer to the output layer, and an activation function of the hidden layer is a radial basis RBF function;
assuming the number of hidden layer nodes is s, the p training sample xpThe output from the ith node of the hidden layer (i ═ 1,2, …, s) is:
Figure BDA0003111294830000034
wherein c isiIs the central value of the ith basis function, and is in the same dimension with the input vector; x is the number ofp-ciIs the kernel function center, and σ is the width parameter of the function; | xp-ci||2Is xpAnd ciThe distance of (c).
Further, in the step (6): the RBF neural network algorithm flow mainly comprises the following steps:
1) network initialization, randomly selecting h training samples as clustering centers ci；
2) Clustering the input training sample set by using a K-Means algorithm;
3) readjusting the clustering centers until the new clustering centers are not changed any more;
4) solving for variance
Figure BDA0003111294830000041
In the formula cmaxThe maximum distance of the selected center;
5) calculating the weight of the hidden layer and the output layer
Figure BDA0003111294830000042
6) Output of
Figure BDA0003111294830000043
Training a neural network by taking the feeder load historical data and the influence factors as input sets to obtain a predicted value of each feeder cluster; and the sum of the predicted values of all the feeder line clusters is the total power load predicted value of the whole feeder line load.
Compared with the prior art, the technical scheme provided by the invention has the following beneficial effects:
the present invention is not limited by the size of data, and can be applied to application scenarios in different regions, and the description of the embodiment should not be a limitation of the present invention. The fuzzy c-means clustering method adopted by the invention can more scientifically and effectively divide the feeder clusters to obtain the power utilization characteristics of each feeder, thereby obtaining more refined load division. And the secondary screening is analyzed by using the correlation analysis and the association rule, so that the obtained influence factors are more accurate and scientific, and the result of load prediction is more accurate.
Drawings
FIG. 1 is a flowchart of a load prediction method based on data mining technology according to the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
Acquiring historical load data of each feeder line in the system, and acquiring the load data according to specific requirements in a specific acquisition process, wherein a sampling point is 96 points of acquisition time, and the acquisition time is a typical day (such as a monthly maximum load occurrence day); abnormal values and null values are processed after data are collected, row deletion processing can be performed when the data volume is large, and in addition, normalization processing needs to be performed on the data before fuzzy C-means clustering is performed.
The electricity consumption behavior analysis step in the invention is based on a clustering analysis technology, namely a fuzzy c-means algorithm, and the flow of the clustering algorithm mainly comprises the following steps:
1, initializing a cluster center: randomly selecting C clustering centers;
2) calculating the mass center: continuously and iteratively calculating membership degree muijAnd cluster center cjUntil they reach the optimum.
Wherein degree of membership
Figure BDA0003111294830000051
Each sample data xiAssign to the nearest cluster center, form C clusters, the objective function is as follows:
Figure BDA0003111294830000052
wherein: m is the number of clusters of the cluster; c is the number of the clustering centers appointed in advance; x is the number ofiThe load curve of the ith feeder line is expressed by a vector; cjIs the center of the j cluster; mu.sijThe membership of the jth sample to the center of the class i load curve.
3) Continuously correcting the clustering center by an iteration method until the preset target function precision is met, namely when | | | muij (k+1)-μij (k)If | | < epsilon, stopping iteration, otherwise, returning to the step (2). Wherein, the step 2) determines that a plurality of clustering clusters, namely a plurality of industries, are finally generated, and the clustering centers are more than 8 according to the fine division of the industries.
Constructing an influence factor set: and (3) collecting economic indexes such as a local total production value and a total per-person production value, collecting climate indexes such as a monthly highest temperature, a lowest temperature and precipitation, and collecting social development indexes such as a urbanization rate and a per-person consumption level through local yearbook information. Constructing a complete influence factor set through the indexes; and constructing an input set of each cluster type prediction model by using the dominant factors.
Firstly, through correlation analysis, finding out factors with strong correlation to feeder data in each feeder cluster, and taking a Pearson correlation coefficient as a calculation standard. And performing association rule analysis on the factors with strong correlation and the feeder line load by using a gray association model, finding out the factors with strong correlation, and constructing an input set of each cluster prediction model as a leading factor. The technical steps are as follows:
1) using the clustered feeder load data as a reference sequence X0Each influencing factor sequence is a comparison sequence Xi。
2) And (5) carrying out dimensionless treatment. And processing each data into the next step of operation of descending the same dimension through normalization processing.
3) And calculating the correlation. The pearson correlation coefficient calculation formula is as follows:
Figure BDA0003111294830000053
the correlation strength of each influence factor and the feeder line data can be calculated through the Pearson correlation coefficient, and then the influence factors with the strong correlation are reserved.
And then performing association rule analysis on the strongly relevant factors, wherein the grey association analysis is adopted in the invention. Grey correlation analysis the technical route in the present invention is as follows:
1) and selecting a reference sequence, namely feeder line data, and comparing the sequences, namely the strong correlation factors after correlation analysis and screening.
2) And (5) data transformation. And processing and transforming the data to ensure that the gray correlation analysis is performed under the same dimension.
3) And calculating the association degree. The present invention employs the dune model, which is a typical gray correlation model. The correlation calculation formula is as follows:
Figure BDA0003111294830000061
wherein: xi (X)0(k),Xi(k) Referred to as reference sequence versus comparison sequence
Figure BDA0003111294830000062
And finally, obtaining factors with strong relevance with various clustered feeder line clusters, and taking the factors and various feeder line loads as input training neural networks.
And (3) feeder line load prediction: the invention realizes load prediction by researching a data mining technology of feeder line data by means of a big data technology architecture. The types of the influence factors are comprehensively considered, an influence factor set is constructed, a data mining technology is applied to the analysis of the power consumption behavior, and a foundation is laid for improving the accuracy of the prediction model.
The feeder load prediction mainly comprises the following steps:
1) and (5) initializing the network. Randomly selecting h training samples as clustering centers ci。
2) The input training sample set is clustered using the K-Means algorithm.
3) The cluster centers are readjusted until the new cluster center is no longer changed.
4) And solving the variance.
Figure BDA0003111294830000063
In the formula cmaxThe maximum distance from the center is selected.
5) And calculating the weight values of the hidden layer and the output layer.
Figure BDA0003111294830000064
6) Output of
Figure BDA0003111294830000071
And training the neural network by taking the feeder load historical data and the influence factors as input sets to obtain the predicted value of each feeder cluster. And the sum of the predicted values of all the feeder line clusters is the total power load predicted value of the whole feeder line load. The training of the neural network can be automatically completed by using SPSS Modeler data mining software to complete the step.
An example is listed below.
Taking feeder load of a certain area as an example, typical daily load data of each month in a sampling time range of 2017-01-2020-07 is sampled at an interval of 15min, and 96 points of data are sampled by each feeder every day. Clustering the fuzzy C mean values to finally obtain 18 cluster clusters. Taking cluster 1 and cluster 6 as an example, correlation analysis is performed, and the result is as follows:
Figure BDA0003111294830000072
the grey correlation analysis was performed for strong correlations, with the following results:
Figure BDA0003111294830000073
the factor GDP and the per-capita GDP with the highest association degree of the cluster 1 can be obtained, and the factor GDP and the per-capita GDP with the highest association degree of the cluster 6 can dominate income.
And (4) respectively carrying out RBF neural network training by taking the correlation factors and various clustering clusters as input, and finally obtaining a prediction result, namely the maximum load of 8 months in 2020. The following is a comparison with a traditional prediction method linear regression model and an RBF neural network prediction model method without considering clustering:
Figure BDA0003111294830000074
finally, it should be noted that the above embodiments are only used for illustrating the technical solutions of the present invention and not for limiting the same. It will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined by the appended claims.

Claims (8)

1. A load prediction method based on a data mining technology is characterized by comprising the following steps:
(1) acquiring feeder line data and preprocessing the feeder line data;
(2) analyzing the load characteristics of various feeder lines by using a clustering method in a data mining technology, and analyzing the obtained load of each feeder line;
(3) the power utilization characteristics of the industry to which each type of feeder load belongs are determined according to the load level and the load curve shape of each feeder cluster;
(4) constructing an influence factor set;
(5) analyzing a plurality of factors influencing the load characteristics and each type of feeder line load by using correlation analysis and association rule analysis, and excavating association characteristics between the feeder line load characteristics and the influence factors thereof;
(6) and predicting each feeder line cluster by using the RBF neural network, obtaining a total power load predicted value of the system at the time to be predicted according to the load predicted value of each feeder line cluster at the time to be predicted, and completing load prediction based on the data mining technology.
2. The method for load prediction based on data mining technology according to claim 1, wherein the step (2) specifically comprises the following steps:
21) initializing a clustering center: randomly selecting C clustering centers;
22) calculating the mass center: continuously and iteratively calculating membership degree muijAnd cluster center cjUntil they reach the optimum;
wherein degree of membership
Figure FDA0003111294820000011
Each sample data xiAssign to the nearest cluster center, form C clusters, the objective function is as follows:
Figure FDA0003111294820000012
wherein: m is the number of clusters of the cluster; c is the number of the clustering centers appointed in advance; x is the number ofiIs the ith sample; cjIs the center of the j cluster; mu.sijIs a sample xiMembership to a jth cluster center;
23) continuously correcting the clustering center by an iteration method until the preset target function precision is met, namely when | | | muij (k+1)-μij (k)Stopping iteration if | | < epsilon, otherwise returning to the step 2);
24) and carrying out dimensionless normalization on the classified feeder line data, and analyzing the electricity utilization characteristics according to the load curve.
3. The load prediction method based on the data mining technology as claimed in claim 1, wherein the set of influence factors of the feeder load in the step (4) includes an economic development index, a social resident development index and a climate index, the economic development index includes GDP, a first industrial production total value, a second industrial production total value, a third industrial production total value and an industrial production total value above a scale, and the social resident development index includes a regional total number, a per capita domestic production total value, a resident consumption price index and a urbanization rate.
4. The method for load prediction based on data mining technology according to claim 1, wherein the step (5) specifically comprises the following steps:
51) after the influence factor set is constructed, judging the correlation between each influence factor and the clustered feeder load data by adopting a Pearson correlation coefficient analysis method;
52) performing association rule analysis on the strongly relevant factors by adopting a grey association analysis method;
53) and obtaining factors with strong relevance, and taking the factors and the clustered various feeder loads as input to train the neural network.
5. The method according to claim 4, wherein the step 51) specifically includes the following steps:
1) using the clustered feeder load data as a reference sequence X0Each influencing factor sequence is a comparison sequence Xi；
2) Dimensionless treatment: processing each data into data under the same dimension through normalization processing;
3) and calculating the correlation, wherein the Pearson correlation coefficient calculation formula is as follows:
Figure FDA0003111294820000021
and calculating the correlation strength of each influence factor and feeder line data through a Pearson correlation coefficient, and then keeping the influence factors with the strong correlation.
6. The method according to claim 4, wherein the step 52) specifically includes the following steps:
1) selecting a reference sequence, namely feeder line data, and comparing the sequences, namely the strong correlation factors after correlation analysis and screening;
2) data transformation: processing and transforming the data to ensure that the grey correlation analysis is carried out under the same dimension;
3) calculating the relevance: the Dun model of a typical gray correlation model is adopted, and the correlation calculation formula is as follows:
Figure FDA0003111294820000022
wherein: xi (X)0(k),Xi(k) Referred to as reference sequence versus comparison sequence
Figure FDA0003111294820000023
7. The load prediction method based on data mining technology as claimed in claim 1, wherein in the step (6):
the data flow of the RBF neural network is as follows: training a sample, namely an RBF hidden layer, a weight matrix and an output layer, wherein nonlinear mapping is performed from the input layer to the output layer, and an activation function of the hidden layer is a radial basis RBF function;
assuming the number of hidden layer nodes is s, the p training sample xpThe output from the ith node of the hidden layer (i ═ 1,2, …, s) is:
Figure FDA0003111294820000031
wherein c isiIs the central value of the ith basis function, and is in the same dimension with the input vector; x is the number ofp-ciIs the kernel function center, and σ is the width parameter of the function; | xp-ci||2Is xpAnd ciThe distance of (c).
8. The load prediction method based on data mining technology as claimed in claim 1, wherein in the step (6): the RBF neural network algorithm flow mainly comprises the following steps:
1) network initialization, randomly selecting h training samples as clustering centers ci；
2) Clustering the input training sample set by using a K-Means algorithm;
3) readjusting the clustering centers until the new clustering centers are not changed any more;
4) solving for variance
Figure FDA0003111294820000032
In the formula cmaxThe maximum distance of the selected center;
5) calculating the weight of the hidden layer and the output layer
Figure FDA0003111294820000033
6) Output of
Figure FDA0003111294820000034
Training a neural network by taking the feeder load historical data and the influence factors as input sets to obtain a predicted value of each feeder cluster; and the sum of the predicted values of all the feeder line clusters is the total power load predicted value of the whole feeder line load.
CN202110651314.8A 2021-06-10 2021-06-10 Load prediction method based on data mining technology Active CN113393028B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110651314.8A CN113393028B (en) 2021-06-10 2021-06-10 Load prediction method based on data mining technology

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110651314.8A CN113393028B (en) 2021-06-10 2021-06-10 Load prediction method based on data mining technology

Publications (2)

Publication Number Publication Date
CN113393028A true CN113393028A (en) 2021-09-14
CN113393028B CN113393028B (en) 2022-08-19

Family

ID=77620452

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110651314.8A Active CN113393028B (en) 2021-06-10 2021-06-10 Load prediction method based on data mining technology

Country Status (1)

Country Link
CN (1) CN113393028B (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113872204A (en) * 2021-12-03 2021-12-31 全球能源互联网研究院有限公司 Power load property determination method and device based on power grid diagram topological calculation
CN114781685A (en) * 2022-03-17 2022-07-22 广西电网有限责任公司 Big user power load prediction method and system based on big data mining technology
CN115829272A (en) * 2022-12-08 2023-03-21 国网江苏省电力有限公司南通供电分公司 Method for extracting key influence factors of industry electric quantity demand

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106485262A (en) * 2016-09-09 2017-03-08 国网山西省电力公司晋城供电公司 A kind of bus load Forecasting Methodology
CN110689195A (en) * 2019-09-26 2020-01-14 云南电网有限责任公司电力科学研究院 Power daily load prediction method
CN111784019A (en) * 2019-12-26 2020-10-16 国网北京市电力公司 Power load processing method and device
CN111815054A (en) * 2020-03-31 2020-10-23 浙江大学 Industrial steam heat supply network short-term load prediction method based on big data
CN112561156A (en) * 2020-12-11 2021-03-26 国网江苏省电力有限公司南通供电分公司 Short-term power load prediction method based on user load mode classification

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106485262A (en) * 2016-09-09 2017-03-08 国网山西省电力公司晋城供电公司 A kind of bus load Forecasting Methodology
CN110689195A (en) * 2019-09-26 2020-01-14 云南电网有限责任公司电力科学研究院 Power daily load prediction method
CN111784019A (en) * 2019-12-26 2020-10-16 国网北京市电力公司 Power load processing method and device
CN111815054A (en) * 2020-03-31 2020-10-23 浙江大学 Industrial steam heat supply network short-term load prediction method based on big data
CN112561156A (en) * 2020-12-11 2021-03-26 国网江苏省电力有限公司南通供电分公司 Short-term power load prediction method based on user load mode classification

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113872204A (en) * 2021-12-03 2021-12-31 全球能源互联网研究院有限公司 Power load property determination method and device based on power grid diagram topological calculation
CN114781685A (en) * 2022-03-17 2022-07-22 广西电网有限责任公司 Big user power load prediction method and system based on big data mining technology
CN114781685B (en) * 2022-03-17 2024-01-09 广西电网有限责任公司 Large user electricity load prediction method and system based on big data mining technology
CN115829272A (en) * 2022-12-08 2023-03-21 国网江苏省电力有限公司南通供电分公司 Method for extracting key influence factors of industry electric quantity demand
CN115829272B (en) * 2022-12-08 2023-07-28 国网江苏省电力有限公司南通供电分公司 Method for extracting key influence factors of electric quantity demand in industry

Also Published As

Publication number Publication date
CN113393028B (en) 2022-08-19

Similar Documents

Publication Publication Date Title
CN113393028B (en) Load prediction method based on data mining technology
CN111199016A (en) DTW-based improved K-means daily load curve clustering method
CN109461025B (en) Electric energy substitution potential customer prediction method based on machine learning
Ozkan et al. Data mining-based upscaling approach for regional wind power forecasting: Regional statistical hybrid wind power forecast technique (RegionalSHWIP)
CN110163429B (en) Short-term load prediction method based on similarity day optimization screening
CN110674993A (en) User load short-term prediction method and device
Zeng et al. A learning framework based on weighted knowledge transfer for holiday load forecasting
CN111461921B (en) Load modeling typical user database updating method based on machine learning
CN111539657B (en) Typical power industry load characteristic classification and synthesis method combined with user daily electricity quantity curve
CN111582548A (en) Power load prediction method based on multivariate user behavior portrait
CN112884590A (en) Power grid enterprise financing decision method based on machine learning algorithm
CN113255900A (en) Impulse load prediction method considering improved spectral clustering and Bi-LSTM neural network
CN115545333A (en) Method for predicting load curve of multi-load daily-type power distribution network
CN116169670A (en) Short-term non-resident load prediction method and system based on improved neural network
Müller et al. Electrical load forecasting in disaggregated levels using fuzzy ARTMAP artificial neural network and noise removal by singular spectrum analysis
CN111126499A (en) Secondary clustering-based power consumption behavior pattern classification method
CN109389517B (en) Analysis method and device for quantifying line loss influence factors
CN116826745B (en) Layered and partitioned short-term load prediction method and system in power system background
Ilich et al. Modeling of employment structural transformations
Wang et al. Cloud computing and extreme learning machine for a distributed energy consumption forecasting in equipment-manufacturing enterprises
CN116307250A (en) Short-term load prediction method and system based on typical daily feature selection
CN107464015B (en) Short-term load with fusing mechanism and electric quantity prediction method based on electricity utilization behavior
Karimi et al. Analyzing the results of buildings energy audit by using grey incidence analysis
CN114091904A (en) Enterprise migration park recruitment analysis method based on artificial intelligence algorithm
CN114372835A (en) Comprehensive energy service potential customer identification method, system and computer equipment

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant